OpenAI to Retire Classic ChatGPT Voice Modes and Transition to GPT-Live
The company announced the sunsetting of older audio versions to focus entirely on its newest conversational model.
OpenAI has announced that it will discontinue both the Advanced Voice Mode and Standard Voice for ChatGPT on December 11, 2026. This transition marks the permanent shift toward GPT-Live, the company's latest voice model which was initially launched in July.
According to the organization, the decision follows a significant increase in user adoption. Since the rollout of GPT-Live, the volume of individuals communicating with ChatGPT through audio has nearly doubled, with the vast majority of active users already migrating to the newer system.
By phasing out the older configurations, OpenAI aims to concentrate its development resources on a single, unified voice experience. The company stated that GPT-Live incorporates key learnings from the previous iterations, resulting in improved response depth, better contextual continuity, and enhanced personalization.
GPT-Live introduces several technical improvements over past models, including the ability to listen and speak simultaneously. The newer architecture allows the system to react to interruptions, use conversational interjections like "mhmm" and "yeah," and pause appropriately while users gather their thoughts.
In contrast, older options carried noticeable limitations. The Advanced Voice Mode operated in rigid conversational turns, requiring the system to wait until a user completely finished speaking before processing a reply. Meanwhile, the Standard Voice option chained together three separate systems to convert speech to text, process the language, and then convert text back into audio, which lengthened response times and occasionally caused conversational nuances to be lost.
OpenAI confirmed that previous user chats conducted with the older voice modes will not be lost. Conversations will remain accessible within the ChatGPT history and can be resumed through text or voice, with ongoing voice interactions defaulting to the GPT-Live system.
The official transition period begins on November 10, 2026, when users will start receiving in-app notifications regarding the upcoming retirement. Both legacy modes will remain available until December 11, after which they will be permanently removed from settings, and any remaining users will be automatically migrated.
Certain features, such as live video and screen sharing, will temporarily continue utilizing the Advanced Voice framework until native support is built directly into GPT-Live. OpenAI has not yet announced a specific timeline for when those native capabilities will arrive.
Sources
- Como proteger o celular Android contra roubo: 8 ajustes e o que fazer tekimobile.com
- OpenAI vai aposentar vozes clássicas e Modo Avançado do ChatGPT canaltech.com.br